Run 4. Outer Step 74. Inner Step 0. Peers 97.
Browse files- config.json +44 -44
- inner_optimizer.pt +1 -1
- model.safetensors +1 -1
- outer_optimizer.pt +1 -1
config.json
CHANGED
@@ -3,7 +3,7 @@
|
|
3 |
"activation_function": "gelu_new",
|
4 |
"all_reduce_scores": {
|
5 |
"0": "NON_PARTICIPATING",
|
6 |
-
"1": "
|
7 |
"10": "NON_PARTICIPATING",
|
8 |
"100": "NON_PARTICIPATING",
|
9 |
"101": "NON_PARTICIPATING",
|
@@ -12,12 +12,12 @@
|
|
12 |
"104": "NON_PARTICIPATING",
|
13 |
"105": "NON_PARTICIPATING",
|
14 |
"106": "NON_PARTICIPATING",
|
15 |
-
"107": "
|
16 |
"108": "NON_PARTICIPATING",
|
17 |
"109": "NON_PARTICIPATING",
|
18 |
"11": "SUCCESS",
|
19 |
"110": "NON_PARTICIPATING",
|
20 |
-
"111": "
|
21 |
"112": "NON_PARTICIPATING",
|
22 |
"113": "NON_PARTICIPATING",
|
23 |
"114": "NON_PARTICIPATING",
|
@@ -26,23 +26,23 @@
|
|
26 |
"117": "NON_PARTICIPATING",
|
27 |
"118": "NON_PARTICIPATING",
|
28 |
"119": "NON_PARTICIPATING",
|
29 |
-
"12": "
|
30 |
"120": "NON_PARTICIPATING",
|
31 |
"121": "NON_PARTICIPATING",
|
32 |
"122": "NON_PARTICIPATING",
|
33 |
"123": "NON_PARTICIPATING",
|
34 |
"124": "NON_PARTICIPATING",
|
35 |
-
"125": "
|
36 |
-
"126": "
|
37 |
"127": "NON_PARTICIPATING",
|
38 |
"128": "NON_PARTICIPATING",
|
39 |
-
"129": "
|
40 |
"13": "SUCCESS",
|
41 |
"130": "NON_PARTICIPATING",
|
42 |
"131": "NON_PARTICIPATING",
|
43 |
"132": "NON_PARTICIPATING",
|
44 |
"133": "NON_PARTICIPATING",
|
45 |
-
"134": "
|
46 |
"135": "NON_PARTICIPATING",
|
47 |
"136": "NON_PARTICIPATING",
|
48 |
"137": "NON_PARTICIPATING",
|
@@ -62,10 +62,10 @@
|
|
62 |
"15": "SUCCESS",
|
63 |
"150": "NON_PARTICIPATING",
|
64 |
"151": "NON_PARTICIPATING",
|
65 |
-
"152": "
|
66 |
-
"153": "
|
67 |
"154": "NON_PARTICIPATING",
|
68 |
-
"155": "
|
69 |
"156": "NON_PARTICIPATING",
|
70 |
"157": "NON_PARTICIPATING",
|
71 |
"158": "NON_PARTICIPATING",
|
@@ -74,40 +74,40 @@
|
|
74 |
"160": "NON_PARTICIPATING",
|
75 |
"161": "NON_PARTICIPATING",
|
76 |
"162": "NON_PARTICIPATING",
|
77 |
-
"163": "
|
78 |
"164": "NON_PARTICIPATING",
|
79 |
"165": "NON_PARTICIPATING",
|
80 |
-
"166": "
|
81 |
"167": "NON_PARTICIPATING",
|
82 |
"168": "NON_PARTICIPATING",
|
83 |
"169": "SUCCESS",
|
84 |
-
"17": "
|
85 |
"170": "NON_PARTICIPATING",
|
86 |
-
"171": "
|
87 |
"172": "NON_PARTICIPATING",
|
88 |
"173": "NON_PARTICIPATING",
|
89 |
"174": "NON_PARTICIPATING",
|
90 |
"175": "NON_PARTICIPATING",
|
91 |
"176": "NON_PARTICIPATING",
|
92 |
-
"177": "
|
93 |
"178": "NON_PARTICIPATING",
|
94 |
"179": "NON_PARTICIPATING",
|
95 |
"18": "SUCCESS",
|
96 |
"180": "NON_PARTICIPATING",
|
97 |
"181": "NON_PARTICIPATING",
|
98 |
"182": "NON_PARTICIPATING",
|
99 |
-
"183": "
|
100 |
"184": "NON_PARTICIPATING",
|
101 |
"185": "NON_PARTICIPATING",
|
102 |
"186": "NON_PARTICIPATING",
|
103 |
"187": "NON_PARTICIPATING",
|
104 |
"188": "NON_PARTICIPATING",
|
105 |
"189": "NON_PARTICIPATING",
|
106 |
-
"19": "
|
107 |
"190": "NON_PARTICIPATING",
|
108 |
"191": "NON_PARTICIPATING",
|
109 |
"192": "NON_PARTICIPATING",
|
110 |
-
"193": "
|
111 |
"194": "NON_PARTICIPATING",
|
112 |
"195": "NON_PARTICIPATING",
|
113 |
"196": "NON_PARTICIPATING",
|
@@ -134,7 +134,7 @@
|
|
134 |
"214": "NON_PARTICIPATING",
|
135 |
"215": "NON_PARTICIPATING",
|
136 |
"216": "NON_PARTICIPATING",
|
137 |
-
"217": "
|
138 |
"218": "NON_PARTICIPATING",
|
139 |
"219": "NON_PARTICIPATING",
|
140 |
"22": "SUCCESS",
|
@@ -148,10 +148,10 @@
|
|
148 |
"227": "NON_PARTICIPATING",
|
149 |
"228": "NON_PARTICIPATING",
|
150 |
"229": "NON_PARTICIPATING",
|
151 |
-
"23": "
|
152 |
"230": "NON_PARTICIPATING",
|
153 |
"231": "NON_PARTICIPATING",
|
154 |
-
"232": "
|
155 |
"233": "NON_PARTICIPATING",
|
156 |
"234": "NON_PARTICIPATING",
|
157 |
"235": "NON_PARTICIPATING",
|
@@ -159,19 +159,19 @@
|
|
159 |
"237": "NON_PARTICIPATING",
|
160 |
"238": "NON_PARTICIPATING",
|
161 |
"239": "NON_PARTICIPATING",
|
162 |
-
"24": "
|
163 |
-
"240": "
|
164 |
"241": "NON_PARTICIPATING",
|
165 |
"242": "NON_PARTICIPATING",
|
166 |
-
"243": "
|
167 |
"244": "NON_PARTICIPATING",
|
168 |
"245": "NON_PARTICIPATING",
|
169 |
"246": "NON_PARTICIPATING",
|
170 |
"247": "NON_PARTICIPATING",
|
171 |
-
"248": "
|
172 |
"249": "NON_PARTICIPATING",
|
173 |
"25": "SUCCESS",
|
174 |
-
"250": "
|
175 |
"251": "NON_PARTICIPATING",
|
176 |
"252": "NON_PARTICIPATING",
|
177 |
"253": "NON_PARTICIPATING",
|
@@ -190,45 +190,45 @@
|
|
190 |
"35": "SUCCESS",
|
191 |
"36": "SUCCESS",
|
192 |
"37": "SUCCESS",
|
193 |
-
"38": "
|
194 |
"39": "SUCCESS",
|
195 |
"4": "SUCCESS",
|
196 |
"40": "SUCCESS",
|
197 |
-
"41": "
|
198 |
"42": "SUCCESS",
|
199 |
"43": "SUCCESS",
|
200 |
"44": "SUCCESS",
|
201 |
"45": "SUCCESS",
|
202 |
"46": "NON_PARTICIPATING",
|
203 |
-
"47": "
|
204 |
"48": "NON_PARTICIPATING",
|
205 |
"49": "SUCCESS",
|
206 |
"5": "SUCCESS",
|
207 |
-
"50": "
|
208 |
"51": "SUCCESS",
|
209 |
"52": "SUCCESS",
|
210 |
"53": "SUCCESS",
|
211 |
"54": "SUCCESS",
|
212 |
-
"55": "
|
213 |
"56": "SUCCESS",
|
214 |
"57": "NON_PARTICIPATING",
|
215 |
"58": "NON_PARTICIPATING",
|
216 |
"59": "NON_PARTICIPATING",
|
217 |
"6": "SUCCESS",
|
218 |
"60": "SUCCESS",
|
219 |
-
"61": "
|
220 |
"62": "NON_PARTICIPATING",
|
221 |
"63": "SUCCESS",
|
222 |
"64": "SUCCESS",
|
223 |
-
"65": "
|
224 |
"66": "SUCCESS",
|
225 |
"67": "SUCCESS",
|
226 |
"68": "SUCCESS",
|
227 |
"69": "SUCCESS",
|
228 |
"7": "SUCCESS",
|
229 |
-
"70": "
|
230 |
"71": "SUCCESS",
|
231 |
-
"72": "
|
232 |
"73": "SUCCESS",
|
233 |
"74": "SUCCESS",
|
234 |
"75": "SUCCESS",
|
@@ -237,27 +237,27 @@
|
|
237 |
"78": "SUCCESS",
|
238 |
"79": "NON_PARTICIPATING",
|
239 |
"8": "NON_PARTICIPATING",
|
240 |
-
"80": "
|
241 |
"81": "SUCCESS",
|
242 |
"82": "SUCCESS",
|
243 |
-
"83": "
|
244 |
"84": "SUCCESS",
|
245 |
"85": "NON_PARTICIPATING",
|
246 |
"86": "NON_PARTICIPATING",
|
247 |
"87": "SUCCESS",
|
248 |
"88": "SUCCESS",
|
249 |
-
"89": "
|
250 |
"9": "SUCCESS",
|
251 |
-
"90": "
|
252 |
"91": "SUCCESS",
|
253 |
"92": "SUCCESS",
|
254 |
"93": "NON_PARTICIPATING",
|
255 |
"94": "SUCCESS",
|
256 |
-
"95": "
|
257 |
"96": "SUCCESS",
|
258 |
-
"97": "
|
259 |
"98": "SUCCESS",
|
260 |
-
"99": "
|
261 |
},
|
262 |
"architectures": [
|
263 |
"GPTOptim"
|
@@ -275,7 +275,7 @@
|
|
275 |
"initializer_range": 0.02,
|
276 |
"inner_step": 0,
|
277 |
"inner_steps": 0,
|
278 |
-
"last_allreduce_block":
|
279 |
"layer_norm_epsilon": 1e-05,
|
280 |
"model_type": "gpt_optimized",
|
281 |
"n_embd": 1280,
|
|
|
3 |
"activation_function": "gelu_new",
|
4 |
"all_reduce_scores": {
|
5 |
"0": "NON_PARTICIPATING",
|
6 |
+
"1": "SUCCESS",
|
7 |
"10": "NON_PARTICIPATING",
|
8 |
"100": "NON_PARTICIPATING",
|
9 |
"101": "NON_PARTICIPATING",
|
|
|
12 |
"104": "NON_PARTICIPATING",
|
13 |
"105": "NON_PARTICIPATING",
|
14 |
"106": "NON_PARTICIPATING",
|
15 |
+
"107": "SUCCESS",
|
16 |
"108": "NON_PARTICIPATING",
|
17 |
"109": "NON_PARTICIPATING",
|
18 |
"11": "SUCCESS",
|
19 |
"110": "NON_PARTICIPATING",
|
20 |
+
"111": "SUCCESS",
|
21 |
"112": "NON_PARTICIPATING",
|
22 |
"113": "NON_PARTICIPATING",
|
23 |
"114": "NON_PARTICIPATING",
|
|
|
26 |
"117": "NON_PARTICIPATING",
|
27 |
"118": "NON_PARTICIPATING",
|
28 |
"119": "NON_PARTICIPATING",
|
29 |
+
"12": "SUCCESS",
|
30 |
"120": "NON_PARTICIPATING",
|
31 |
"121": "NON_PARTICIPATING",
|
32 |
"122": "NON_PARTICIPATING",
|
33 |
"123": "NON_PARTICIPATING",
|
34 |
"124": "NON_PARTICIPATING",
|
35 |
+
"125": "SUCCESS",
|
36 |
+
"126": "SUCCESS",
|
37 |
"127": "NON_PARTICIPATING",
|
38 |
"128": "NON_PARTICIPATING",
|
39 |
+
"129": "SUCCESS",
|
40 |
"13": "SUCCESS",
|
41 |
"130": "NON_PARTICIPATING",
|
42 |
"131": "NON_PARTICIPATING",
|
43 |
"132": "NON_PARTICIPATING",
|
44 |
"133": "NON_PARTICIPATING",
|
45 |
+
"134": "NON_PARTICIPATING",
|
46 |
"135": "NON_PARTICIPATING",
|
47 |
"136": "NON_PARTICIPATING",
|
48 |
"137": "NON_PARTICIPATING",
|
|
|
62 |
"15": "SUCCESS",
|
63 |
"150": "NON_PARTICIPATING",
|
64 |
"151": "NON_PARTICIPATING",
|
65 |
+
"152": "NON_PARTICIPATING",
|
66 |
+
"153": "NON_PARTICIPATING",
|
67 |
"154": "NON_PARTICIPATING",
|
68 |
+
"155": "NON_PARTICIPATING",
|
69 |
"156": "NON_PARTICIPATING",
|
70 |
"157": "NON_PARTICIPATING",
|
71 |
"158": "NON_PARTICIPATING",
|
|
|
74 |
"160": "NON_PARTICIPATING",
|
75 |
"161": "NON_PARTICIPATING",
|
76 |
"162": "NON_PARTICIPATING",
|
77 |
+
"163": "SUCCESS",
|
78 |
"164": "NON_PARTICIPATING",
|
79 |
"165": "NON_PARTICIPATING",
|
80 |
+
"166": "NON_PARTICIPATING",
|
81 |
"167": "NON_PARTICIPATING",
|
82 |
"168": "NON_PARTICIPATING",
|
83 |
"169": "SUCCESS",
|
84 |
+
"17": "SUCCESS",
|
85 |
"170": "NON_PARTICIPATING",
|
86 |
+
"171": "SUCCESS",
|
87 |
"172": "NON_PARTICIPATING",
|
88 |
"173": "NON_PARTICIPATING",
|
89 |
"174": "NON_PARTICIPATING",
|
90 |
"175": "NON_PARTICIPATING",
|
91 |
"176": "NON_PARTICIPATING",
|
92 |
+
"177": "NON_PARTICIPATING",
|
93 |
"178": "NON_PARTICIPATING",
|
94 |
"179": "NON_PARTICIPATING",
|
95 |
"18": "SUCCESS",
|
96 |
"180": "NON_PARTICIPATING",
|
97 |
"181": "NON_PARTICIPATING",
|
98 |
"182": "NON_PARTICIPATING",
|
99 |
+
"183": "SUCCESS",
|
100 |
"184": "NON_PARTICIPATING",
|
101 |
"185": "NON_PARTICIPATING",
|
102 |
"186": "NON_PARTICIPATING",
|
103 |
"187": "NON_PARTICIPATING",
|
104 |
"188": "NON_PARTICIPATING",
|
105 |
"189": "NON_PARTICIPATING",
|
106 |
+
"19": "SUCCESS",
|
107 |
"190": "NON_PARTICIPATING",
|
108 |
"191": "NON_PARTICIPATING",
|
109 |
"192": "NON_PARTICIPATING",
|
110 |
+
"193": "SUCCESS",
|
111 |
"194": "NON_PARTICIPATING",
|
112 |
"195": "NON_PARTICIPATING",
|
113 |
"196": "NON_PARTICIPATING",
|
|
|
134 |
"214": "NON_PARTICIPATING",
|
135 |
"215": "NON_PARTICIPATING",
|
136 |
"216": "NON_PARTICIPATING",
|
137 |
+
"217": "SUCCESS",
|
138 |
"218": "NON_PARTICIPATING",
|
139 |
"219": "NON_PARTICIPATING",
|
140 |
"22": "SUCCESS",
|
|
|
148 |
"227": "NON_PARTICIPATING",
|
149 |
"228": "NON_PARTICIPATING",
|
150 |
"229": "NON_PARTICIPATING",
|
151 |
+
"23": "SUCCESS",
|
152 |
"230": "NON_PARTICIPATING",
|
153 |
"231": "NON_PARTICIPATING",
|
154 |
+
"232": "NON_PARTICIPATING",
|
155 |
"233": "NON_PARTICIPATING",
|
156 |
"234": "NON_PARTICIPATING",
|
157 |
"235": "NON_PARTICIPATING",
|
|
|
159 |
"237": "NON_PARTICIPATING",
|
160 |
"238": "NON_PARTICIPATING",
|
161 |
"239": "NON_PARTICIPATING",
|
162 |
+
"24": "SUCCESS",
|
163 |
+
"240": "NON_PARTICIPATING",
|
164 |
"241": "NON_PARTICIPATING",
|
165 |
"242": "NON_PARTICIPATING",
|
166 |
+
"243": "NON_PARTICIPATING",
|
167 |
"244": "NON_PARTICIPATING",
|
168 |
"245": "NON_PARTICIPATING",
|
169 |
"246": "NON_PARTICIPATING",
|
170 |
"247": "NON_PARTICIPATING",
|
171 |
+
"248": "NON_PARTICIPATING",
|
172 |
"249": "NON_PARTICIPATING",
|
173 |
"25": "SUCCESS",
|
174 |
+
"250": "SUCCESS",
|
175 |
"251": "NON_PARTICIPATING",
|
176 |
"252": "NON_PARTICIPATING",
|
177 |
"253": "NON_PARTICIPATING",
|
|
|
190 |
"35": "SUCCESS",
|
191 |
"36": "SUCCESS",
|
192 |
"37": "SUCCESS",
|
193 |
+
"38": "SUCCESS",
|
194 |
"39": "SUCCESS",
|
195 |
"4": "SUCCESS",
|
196 |
"40": "SUCCESS",
|
197 |
+
"41": "SUCCESS",
|
198 |
"42": "SUCCESS",
|
199 |
"43": "SUCCESS",
|
200 |
"44": "SUCCESS",
|
201 |
"45": "SUCCESS",
|
202 |
"46": "NON_PARTICIPATING",
|
203 |
+
"47": "SUCCESS",
|
204 |
"48": "NON_PARTICIPATING",
|
205 |
"49": "SUCCESS",
|
206 |
"5": "SUCCESS",
|
207 |
+
"50": "SUCCESS",
|
208 |
"51": "SUCCESS",
|
209 |
"52": "SUCCESS",
|
210 |
"53": "SUCCESS",
|
211 |
"54": "SUCCESS",
|
212 |
+
"55": "SUCCESS",
|
213 |
"56": "SUCCESS",
|
214 |
"57": "NON_PARTICIPATING",
|
215 |
"58": "NON_PARTICIPATING",
|
216 |
"59": "NON_PARTICIPATING",
|
217 |
"6": "SUCCESS",
|
218 |
"60": "SUCCESS",
|
219 |
+
"61": "SUCCESS",
|
220 |
"62": "NON_PARTICIPATING",
|
221 |
"63": "SUCCESS",
|
222 |
"64": "SUCCESS",
|
223 |
+
"65": "SUCCESS",
|
224 |
"66": "SUCCESS",
|
225 |
"67": "SUCCESS",
|
226 |
"68": "SUCCESS",
|
227 |
"69": "SUCCESS",
|
228 |
"7": "SUCCESS",
|
229 |
+
"70": "SUCCESS",
|
230 |
"71": "SUCCESS",
|
231 |
+
"72": "SUCCESS",
|
232 |
"73": "SUCCESS",
|
233 |
"74": "SUCCESS",
|
234 |
"75": "SUCCESS",
|
|
|
237 |
"78": "SUCCESS",
|
238 |
"79": "NON_PARTICIPATING",
|
239 |
"8": "NON_PARTICIPATING",
|
240 |
+
"80": "SUCCESS",
|
241 |
"81": "SUCCESS",
|
242 |
"82": "SUCCESS",
|
243 |
+
"83": "SUCCESS",
|
244 |
"84": "SUCCESS",
|
245 |
"85": "NON_PARTICIPATING",
|
246 |
"86": "NON_PARTICIPATING",
|
247 |
"87": "SUCCESS",
|
248 |
"88": "SUCCESS",
|
249 |
+
"89": "SUCCESS",
|
250 |
"9": "SUCCESS",
|
251 |
+
"90": "SUCCESS",
|
252 |
"91": "SUCCESS",
|
253 |
"92": "SUCCESS",
|
254 |
"93": "NON_PARTICIPATING",
|
255 |
"94": "SUCCESS",
|
256 |
+
"95": "SUCCESS",
|
257 |
"96": "SUCCESS",
|
258 |
+
"97": "SUCCESS",
|
259 |
"98": "SUCCESS",
|
260 |
+
"99": "SUCCESS"
|
261 |
},
|
262 |
"architectures": [
|
263 |
"GPTOptim"
|
|
|
275 |
"initializer_range": 0.02,
|
276 |
"inner_step": 0,
|
277 |
"inner_steps": 0,
|
278 |
+
"last_allreduce_block": 5661125,
|
279 |
"layer_norm_epsilon": 1e-05,
|
280 |
"model_type": "gpt_optimized",
|
281 |
"n_embd": 1280,
|
inner_optimizer.pt
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 8081782026
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:6b85f7f1d4e749ada7fb24ca96e82be49ab851ebb34a6c89bfd745b4963d2f00
|
3 |
size 8081782026
|
model.safetensors
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 4040701744
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:0d77f4b4518edeedaa61ef8697e57157b385bb05b3764fdc1e1a4397aabc35e1
|
3 |
size 4040701744
|
outer_optimizer.pt
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 4040805354
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:bfed9d96d6466d30fd441e5b1e0883b8f602878c21213761703932bd7ef1559f
|
3 |
size 4040805354
|